Almost 75% of HF patients, a group of about 5.1 million Americans who have poor function and high levels of morbidity and mortality, report poor sleep. As many as 25-56% of HF patients report chronic insomnia (difficulty initiating or maintaining sleep or waking early in the morning, with non-restorative sleep that persists for at least a month). Chronic insomnia may contribute to the development and exacerbation of HF and incident mortality. It is also associated with common and disabling symptoms (fatigue, dyspnea, anxiety, depression, excessive daytime sleepiness, and pain) and decrements in objective and subjective functional performance.10 However, insomnia is under-diagnosed and under-treated in this population. The purposes of the proposed RCT are to evaluate the sustained effects of CBT-I, compared with HF self-management education (attention control), on insomnia severity, sleep characteristics, daytime symptoms, and functional performance over twelve months among patients who have stable chronic HF and chronic insomnia and receive evidence-based HF disease management. We will also evaluate the cost-effectiveness of CBT-I compared with the attention- control condition and explore the effects of CBT-I on event-free survival. We will address the following specific aims (*primary outcomes): (1) Test the sustained effects (baseline - 2 weeks, 6, 9, 12 months) of CBT-I provided in 4 group sessions over 8 weeks, compared with HF self-management education (attention control condition), on: (1a) *insomnia severity and self-reported and actigraph-recorded sleep characteristics (*sleep quality, *sleep efficiency, sleep latency, and duration); (1b) symptoms (*fatigue, anxiety, depression, pain, sleepiness, sleep-related impairment), and psychomotor vigilance (PVT); and (1c) symptom clusters [membership in clusters characterized by severity of specific symptoms; transition between clusters over time]; (2) Test the sustained effects of CBT-I on self-reported and objective functional performance; and (3) Examine the cost-effectiveness of CBT-I. Exploratory aim: We will explore the effects of CBT-I on event-free survival. A total of 200 patients will be randomized to 4 bi-weekly sessions of group CBT-I or an attention control condition consisting of HF self-management education. Wrist actigraph measures of sleep, diaries, psychomotor vigilance and 6 minute walk test distance, and self-report measures of insomnia, sleep, symptoms, and functional performance will be obtained at baseline and follow-up. Data analysis will consist of mixed effects models, latent transition analysis, stochastic cost effectiveness analysis, and survival analysis.